This is a perennial problem in any kind of government service, where you're asking yourself, “What does this bureaucrat do in terms of taking an idea and turning it into output?” but that's true of everyone. How do you measure this in the health sector, for example? It's very difficult to measure outcomes in government services, and rightly so in some cases.
Where I would go in terms of value for money is in the budget process—for example, there are ways of asking yourself when you have a clear fiscal anchor of a return to balanced budget in, say, a few years. Rather than a much more high-level question of an anchor that's based on debt to GDP, which depends on multiple factors, including the economy, when you have a balanced budget requirement in a few years, you have a very clear set of metrics that say, “This spending is not going to fit within our set criteria.” If we have that kind of outlook of when we're trying to return to a balanced budget, it will help government really focus its mind on what kind of spending makes the most sense.